Dataset Information

A stacking-based model for predicting 30-day all-cause hospital readmissions of patients with acute myocardial infarction.

ABSTRACT:

Background

Acute myocardial infarction (AMI) is a serious cardiovascular disease, followed by a high readmission rate within 30-days of discharge. Accurate prediction of AMI readmission is a crucial way to identify the high-risk group and optimize the distribution of medical resources.

Methods

In this study, we propose a stacking-based model to predict the risk of 30-day unplanned all-cause hospital readmissions for AMI patients based on clinical data. Firstly, we conducted an under-sampling method of neighborhood cleaning rule (NCR) to alleviate the class imbalance and then utilized a feature selection method of SelectFromModel (SFM) to select effective features. Secondly, we adopted a self-adaptive approach to select base classifiers from eight candidate models according to their performances in datasets. Finally, we constructed a three-layer stacking model in which layer 1 and layer 2 were base-layer and level 3 was meta-layer. The predictions of the base-layer were used to train the meta-layer in order to make the final forecast.

Results

The results show that the proposed model exhibits the highest AUC (0.720), which is higher than that of decision tree (0.681), support vector machine (0.707), random forest (0.701), extra trees (0.709), adaBoost (0.702), bootstrap aggregating (0.704), gradient boosting decision tree (0.710) and extreme gradient enhancement (0.713).

Conclusion

It is evident that our model could effectively predict the risk of 30-day all cause hospital readmissions for AMI patients and provide decision support for the administration.

SUBMITTER: Zhang Z

PROVIDER: S-EPMC7734833 | biostudies-literature | 2020 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

A stacking-based model for predicting 30-day all-cause hospital readmissions of patients with acute myocardial infarction.

Zhang Zhen Z Qiu Hang H Li Weihao W Chen Yucheng Y

BMC medical informatics and decision making 20201214 1

<h4>Background</h4>Acute myocardial infarction (AMI) is a serious cardiovascular disease, followed by a high readmission rate within 30-days of discharge. Accurate prediction of AMI readmission is a crucial way to identify the high-risk group and optimize the distribution of medical resources.<h4>Methods</h4>In this study, we propose a stacking-based model to predict the risk of 30-day unplanned all-cause hospital readmissions for AMI patients based on clinical data. Firstly, we conducted an und ...[more]

PMID: 33317534

Similar Datasets

Project description:To better guide strategies intended to reduce high rates of 30-day readmission after hospitalization for heart failure (HF), acute myocardial infarction (MI), or pneumonia, further information is needed about readmission diagnoses, readmission timing, and the relationship of both to patient age, sex, and race.To examine readmission diagnoses and timing among Medicare beneficiaries readmitted within 30 days after hospitalization for HF, acute MI, or pneumonia.We analyzed 2007-2009 Medicare fee-for-service claims data to identify patterns of 30-day readmission by patient demographic characteristics and time after hospitalization for HF, acute MI, or pneumonia. Readmission diagnoses were categorized using an aggregated version of the Centers for Medicare & Medicaid Services' Condition Categories. Readmission timing was determined by day after discharge.We examined the percentage of 30-day readmissions occurring on each day (0-30) after discharge; the most common readmission diagnoses occurring during cumulative periods (days 0-3, 0-7, 0-15, and 0-30) and consecutive periods (days 0-3, 4-7, 8-15, and 16-30) after hospitalization; median time to readmission for common readmission diagnoses; and the relationship between patient demographic characteristics and readmission diagnoses and timing.From 2007 through 2009, we identified 329,308 30-day readmissions after 1,330,157 HF hospitalizations (24.8% readmitted), 108,992 30-day readmissions after 548,834 acute MI hospitalizations (19.9% readmitted), and 214,239 30-day readmissions after 1,168,624 pneumonia hospitalizations (18.3% readmitted). The proportion of patients readmitted for the same condition was 35.2% after the index HF hospitalization, 10.0% after the index acute MI hospitalization, and 22.4% after the index pneumonia hospitalization. Of all readmissions within 30 days of hospitalization, the majority occurred within 15 days of hospitalization: 61.0%, HF cohort; 67.6%, acute MI cohort; and 62.6%, pneumonia cohort. The diverse spectrum of readmission diagnoses was largely similar in both cumulative and consecutive periods after discharge. Median time to 30-day readmission was 12 days for patients initially hospitalized for HF, 10 days for patients initially hospitalized for acute MI, and 12 days for patients initially hospitalized for pneumonia and was comparable across common readmission diagnoses. Neither readmission diagnoses nor timing substantively varied by age, sex, or race.Among Medicare fee-for-service beneficiaries hospitalized for HF, acute MI, or pneumonia, 30-day readmissions were frequent throughout the month after hospitalization and resulted from a similar spectrum of readmission diagnoses regardless of age, sex, race, or time after discharge.

Dataset Information

A stacking-based model for predicting 30-day all-cause hospital readmissions of patients with acute myocardial infarction.

Background

Methods

Results

Conclusion

Publications

A stacking-based model for predicting 30-day all-cause hospital readmissions of patients with acute myocardial infarction.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure